Currently, marijuana (MJ) is the most popular illicit drug, with prevalence studies indicating increasing use among young adults (Johnston et al., 2011). Even so, there are few effective interventions to help MJ users reduce their intake to avoid negative consequences, including MJ dependence. In this revised application, we propose a Stage 1 efficacy study to develop and initially test an innovative intervention to reduce MJ use among young adults who regularly use MJ (> 3 episodes/week). The intervention includes elements from the Marijuana Check-Up (MCU; Stephens et al., 2007), a MI-based brief intervention that has shown promise for reducing MJ use. It also incorporates findings from our ongoing research, which suggest that exercise/physical activity (PA) has potential as a positive alternative to MJ use. Our research also has shown that short (i.e., 10 minute) bouts of moderate or intense exercise reduce craving/urges to use MJ. Exercise interventions have successfully reduced use of licit substances, such as tobacco (e.g., Marcus et al., 2005) and alcohol (e.g., Brown et al., 2009), but have not been adequately tested for MJ use. The two aims of this R34 application are: 1) To develop an intervention that consists of four, 60-minute, in-person sessions composed of MCU content (e.g., personalized feedback, MI) as well as a smartphone application (app) that promotes exercise/physical activity (EA) as an alternative to MJ use. The EA, which will be designed to specifically appeal to young adults, will provide a readily-accessible, flexible, and convenient platform for personalized information and reminders that promote exercise/PA as a positive alternative to MJ use in ongoing daily life. 2) To conduct a pilot/efficacy study of the 4-week MCU+EA intervention vs. a MCU-only control condition. We will use urn randomization to assign emerging/young-adult MJ users (N = 40) to the two conditions. During the 1-week baseline, 4-week intervention phase, and 1-, 3-, and 6-month follow-ups, all participants will use the smartphone app to provide real-time data on MJ-related variables and wear accelerometers to provide PA data. Multilevel modeling will be used to test our hypothesis that the MCU+EA intervention, compared to MCU-only control, will produce greater decreases in quantity and frequency of MJ use (and related MJ problems) at post-intervention and at each follow-up. Multilevel modeling also will be used to explore the real-time data to examine the role of variables such as urge/craving for MJ, social factors, and dosage of PA in the reduction of the quantity and frequency of MJ use. This Stage 1 study is unique and innovative in its development and use of a smartphone app to promote and evaluate exercise/PA as a positive alternative to MJ use in daily life. It includes cutting-edge technology (e.g., accelerometers, smartphone app) for real-time assessments. This research will make significant contributions to the limited knowledge of exercise/PA as a strategy for reducing MJ use and related problems among emerging and young adults.